Micromechanics-based variational phase-field modeling of Brazilian splitting tests

The simulation of the Brazilian splitting test can be challenging due to the presence of stress concentrations near the load points, which can lead to numerical problems and convergence issues. This paper aims to realistically model the Brazilian splitting test on mortar samples under monotonic loading, employing a micromechanics-based variational phase-field formulation. A micromechanics-based phase-field model connects the field variables at the macroscale and the physical dissipative mechanisms at the microcrack level, associating plasticity with the sliding of closed microcracks and damage with the growth of open microcracks. The distinction between these two mechanisms aids crack initiation and helps to overcome convergence issues. To further support convergence, a viscoplasticity regularization is embedded in the micromechanics-based phase field model. First, the homogeneous response of a single volume element is presented to demonstrate the effect of the viscosity parameter, and to provide insight of different failure modes that the model predicts. The model is next calibrated against experimental results. Hereto, Brazilian splitting tests have been performed using Digital Image Correlation technique. The calibrated numerical model predicts shear fracture mechanisms involved in the Brazilian splitting test and this has been further explained by studying the effect of confining pressure on modulated failure modes in the model. The calibrated parameters are then used for other samples with the same material but a different size and boundary conditions to assess the validity of the model and its prediction capabilities. Results show that the numerical failure prediction is consistent with the experiment. Overall, the calibration and comparison of numerical and experimental results confirm the validity and reliability of the simulation model and its ability to predict material behavior and fracture modes under different boundary conditions.